Predicting Transcriptional and Epigenetic Networks in Cancer from Sequencing Data

描述（由申请人提供）：该项目旨在开发生物信息资源，用于处理和整合快速出现的大规模测序数据，用于研究癌症研究中的致癌转录因子（TF）。虽然我们的方法将适用于一般的转录因子，但我们将通过关注小眼相关转录因子（MITF）来开发我们的工具，MITF是一种在黑色素瘤中经常扩增的关键肿瘤蛋白。 MITF 可能是黑色素瘤中研究最深入的 TF，负责将多种信号转化为增殖、存活和侵袭的转录控制。因此，研究 MITF 等致癌 TF 的机制并全面鉴定其直接靶基因仍然是癌症研究中尚未解决的重要问题。基于高通量 DNA 测序的癌症基因组学现在正在快速生成大量遗传和表观遗传数据，这些数据可以共同揭示 MITF 如何发挥黑色素瘤进展的有效调节作用。分析如此庞大的异构数据集经常面临测序失败和缺乏用于整合和解释结果信息的分析方法的挑战。所提出的工具将解决这些紧迫的问题：（1）我们将为 ChIP-seq 和 RNA-seq 数据开发独立于平台的质量控制可视化软件。我们的软件包将自动测试并以图形方式总结数据质量，并建议潜在的故障源； (2) 我们将开发和应用计算工具来发现 MITF 的合作 TF。 TF 结合活性本身通常不足以调节基因表达，这表明协同因子的特定组合至关重要地决定 MITF 转录黑色素瘤中关键癌基因的能力。因此，我们将通过将 ChIP-seq 数据与 DNA 序列分析相结合，以计算方式识别并通过实验验证 MITF 的协同因素； (3) 我们将开发和应用统计方法来推断由 MITF 控制并指导的表观遗传变化，从而识别破坏 MITF 正常功能的异常表观遗传修饰； （4）由于非编码RNA（ncRNA）和逆转录转座子的异常表达可以严重改变细胞周期、凋亡和增殖，因此我们将鉴定黑色素瘤中的活性ncRNA和逆转录转座子并发现它们的转录调节因子。这些结果将有助于揭示 MITF 在黑色素瘤中的转录和表观遗传网络，并产生适用于其他癌症的宝贵资源。

DESCRIPTION (provided by applicant): This project aims to develop bioinformatic resources for processing and integrating the large-scale sequencing data that are rapidly emerging for studying oncogenic transcription factors (TFs) in cancer research. While our methods will be applicable to general TFs, we will develop our tools by focusing on microphthalmia-associated transcription factor (MITF), a key onco-protein frequently amplified in melanoma. MITF is perhaps the most intensely studied TF in melanoma, being responsible for turning multiple signals into a transcriptional control of proliferation, survival, and invasion. Studying the mechanisms of an oncogenic TF, such as MITF, and comprehensively identifying its direct target genes thus remain important unsolved problems in cancer research. Cancer genomics based on high-throughput DNA sequencing is now rapidly generating enormous amounts of genetic and epigenetic data that can collectively reveal how MITF functions as a potent regulator of melanoma progression. Analyzing such massive heterogeneous datasets is frequently challenged by both sequencing failures and the lack of analysis methods for integrating and interpreting the resulting information. The proposed tools will address these urgent problems: (1) We will develop a stand-alone platform- independent quality control visualization software for ChIP-seq and RNA-seq data. Our software package will automatically test and graphically summarize the quality of data and also suggest potential sources of failure; (2) We will develop and apply computational tools for discovering cooperating TFs of MITF. TF binding activity in itself is often insufficient to regulate gene expression, suggesting that specific combinations of cooperating factors crucially determine MITF's ability to transcribe key oncogenes in melanoma. We will thus computationally identify and experimentally validate cooperating factors of MITF by combining ChIP-seq data with DNA sequence analysis; (3) We will develop and apply statistical methods for inferring the epigenetic changes that are both controlled by and guiding MITF and, as a result, identify aberrant epigenetic modifications that disrupt normal MITF functions; (4) As aberrant expression of non-coding RNAs (ncRNAs) and retrotransposons can critically alter cell cycle, apoptosis and proliferation, we will identify active ncRNAs and retrotransposons in melanoma and discover their transcriptional regulators. These results will help reveal the transcriptional and epigenetic network of MITF in melanoma and produce valuable resources applicable to other cancers.